| | Design : Test Plans | 
|
|
In the desired outcomes section of the analysis you should have made a list of the things that you want your system to do. Before creating the system the last thing you need to do is to come up with a test plan explaining how you will check whether or not your system meets these requirements.
For each requirement that the system should meet you need to describe how you will test if the system does in fact meet the requirement. Each of the tests that you might perform falls into one of two categories :
An absolute test is a test to which there is a definite pass or fail answer. For example, if you have said that the system should produce a list of all of the books that are out on loan from a library you can test if the system will do this by getting the computer to produce the list and manually checking if it contains the right books or not. You should perform absolute tests with three different types of test data : Typical, Extreme and Erroneous. The table below explains what these different types of data are and gives examples of them for dates and percentages.
| Type | Description | Percentage e.g. | Date e.g.
| Typical Data
| Typical, everyday data which your information system should work with.
| 23, 67
| 23/4/1998, 2/12/2001
| Extreme Data
| Data which is acceptable but is on the edge of the boundaries for which your information system should work.
| 0,100
| 30/6/2001, 29/2/2000
| Erroneous Data
| Data which should not work with you information system. A good information system will
display an error message when erroneous data is entered and will not crash.
| -1, 101, 150
| 31/6/2001, 29/2/2001
| |
A subjective test is a test in which a judgement must be made by a human about the computer system. e.g. Is the logo that was designed appealing, or is the data collection form easy to understand and complete correctly ? Often there is no definite answer to these questions. The best thing you can do is ask as many people as possible to give you their opinions or to try the system out.
Sometimes you may not be able to think of a way to test a few of the requirements. Or it may be too complicated to test if a requirement is met. It is enough to make sure that you test the majority of the system.
Testing Different Types Of Task
Here are some methods you may use to test whether a computer system works :
| Type of Task | Likely Method of Testing |
| DTP / Word Processing / Graphics | Visual inspection to check that final output matches design, ask other peoples' views on the work - do they understand it, does it achieve its purpose, does it contain all of the required information? |
| Spreadsheet | Perform calculations by hand and compare results to computer output results. |
| Database | Perform queries / reports and check that the records that are output are the correct ones. Check that validation checks accept / reject the correct data. |
For each test you are going to carry out you must state these things in your test plan :
It is quite acceptable for your system to fail some tests. If this happens then you should explain why, and if possible correct the problem(s) then retest. Your system may also fail some tests because you have improved it. For example you may change a poster from its design when you create it because you believe the changes amount to an improvement. If this is the case you must explain why it is an improvement.
Testing can not prove that an information system will work all of the time. But the more testing you do the more confident you can be about how well your system works.
GCSE ICT Companion 04 - (C) P Meakin 2004